Environmental Immunochemical Methods - American Chemical Society

can then capture any one of nine specific antibodies which may be introduced into ... In addition to the 0 sample to evaluate binding to the microwell...
0 downloads 0 Views 696KB Size
Chapter 7

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

Development and Application of an EnzymeLinked Immunosorbent Assay Method for the Determination of Multiple Sulfonylurea Herbicides on the Same Microwell Plate Johanne Strahan DuPont Agricultural Products, Experimental Station, Wilmington, DE 19880-0402

A competitive enzyme linked immunosorbent assay (ELISA) was developed and optimized for the simultaneous analysis of multiple DuPont sulfonylurea herbicides (SUs) on the same polyantigen coated microwell plate. As many as 9 different antibodies can be assayed on this type of plate. The same excellent sensitivity, precision, and accuracy observed in the standard quantitative method is also obtained on a polyantigen coated microwell plate. Reagent optimization to increase assay sensitivity led to the development of a polyantigen coating. This ELISA format allows for a very high sample throughput and was used to screen 1500 boxes of Benlate DF fungicide (1313 discrete lots) for nine SUs at an LOD of 5 ppb in the formulated product. Assay Format

A competitive ELISA was developed and optimized for the simultaneous analysis of more than one sulfonylurea herbicide on the same microwell plate. For the assay, multiple portions of the same sample are prepared and one specific antibody is introduced into each portion. The polyantigen coating on the microwell plate can then capture any one of nine specific antibodies which may be introduced into an aliquot (portion) of the sample. Figure 1 is a schematic of the assay format. In step 1, the specific antibody is added to the sample. After preincubation, an aliquot of the sample is pipetted into the precoated wells on the microwell plate. During this next incubation period, the antibody that did not bind to the antigen in the sample, will now bind to the antigen immobilized on the microwell plate. The plate is then washed. In step 3, a second antibody enzyme conjugate is pipetted into the wells. The second antibody binds to the immobilized first antibody. After another wash step, the enzyme substrate is added and a color develops. The intensity of the color is inversely related to the concentration of the antigen. If there is no antigen in the sample, there is maximum color. With increasing amounts of antigen, there is decreasing color. The absorbance can be measured on a microwell plate reader and with the appropriate software, standards and controls, concentrations in samples may be calculated in a quantitative assay or estimated in a screening assay. 0097-6156/96/0646-0065$15.00/0 © 1996 American Chemical Society Van Emon et al.; Environmental Immunochemical Methods ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

66

ENVIRONMENTAL IMMUNOCHEMICAL METHODS

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

1. Incubate sample containing antigen · with specific antibody Y in a tube.

2. Add solution from Step 1. to microplate wells coated with antigen-protein conjugate. Incubate. Antibodies not bound to antigen in sample will bind to antigen on microplate. Wash microplate

Add second antibody-enzyme conjugate E r c | ' Incubate. Second antibody-enzyme binds to antibodies bound to microplate. Wash microplate.

4 Add enzyme substrate. Incubate. Read absorbance. Ο Antigen = maximum color + Antigen = minimum color Figure 1. ELISA (Enzyme Lir

Immunosorbent Assay)

Van Emon et al.; Environmental Immunochemical Methods ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

7. STRAHAN

ELISA Method for Analysis of Multiple SUs

67

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

Assay Components

Polyclonal antibodies specific for the sulfonylurea herbicide (SU) were produced in New Zealand white rabbits by conventional immunization procedures. The immunogen used consisted of the SU hapten conjugated to keyhole limpet hemocyanin (KLH). The resultant antibodies are not purified but left in their natural milieu, the rabbit serum. Antibody tablets are formulated by a DuPont proprietary process. These may contain as little as 1 uL/tablet or as much as 5 uL/tablet of the antisera. They are designed so that the addition of a tablet to a certain amount of buffer results in the desired antibody titer. Coating antigens are prepared by conjugating the SU hapten to ovalbumin using conventional conjugation protocols; i.e., activation of a carboxylic acid derivative of the antigen followed by reaction with the amines of a carrier protein (KLH) which results in a hydrolytically stable amide bond. Second antibody alkaline phosphatase reagent is commercially available. The substrate, p-nitophenyl-phosphate (PNPP) is also commercially available. Microwell plates are availablefromseveral vendors and the best quality plates with optimal protein-adsorptive capacity are used. Plates are coated by pipetting 200 uL/well of the coating antigen mixture in a phosphate buffer, pH 7.4 (PBS). They are coated at ambient temperature and usually left on lab bench overnight and washed in the morning as this fits in with the work flow. Plates are then stored in a zip-lock plastic bag with desiccant at 4°C. Assay Optimization

Antibody titer is determined in a checkerboard format with different coating antigen concentrations going down the plate and different antibody dilutions going across the plate. In addition to the 0 sample to evaluate binding to the microwell plate, another sample is included to evaluate inhibition. After the optimal antibody titer is selected for 50 % inhibition in the middle of the range of sensitivity desired, then a complete standard curve is run with the optimal coating antigen concentration and antibodytiter.In addition, further investigations may be made to attempt to increase sensitivity. For example, Table I shows increased sensitivity with decreasing concentrations of antibody in the assay. Table L Antibody Optimization % Inhibition* 25ppt % Antibody 5ppt Wppt 11 100 6 42 80 13 25 44 60 11 26 40 10 30 56 67 20 29 47

ii

50ppt 52 61 65 72 75

*% Inhibition = ((A-B)/A) χ 100 where A = Absorbance of Negative Control Β = Absorbance of Sample Coating antigen is initially optimized along with the antibody in the checkerboard assay as described. The goal is to reduce the concentration of the coating antigen and hence the antibody without sacrificing the robustness of the

Van Emon et al.; Environmental Immunochemical Methods ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

68

ENVIRONMENTAL IMMUNOCHEMICAL METHODS

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

assay or any of the assay performance specifications, i.e. precision. Table Π shows the effect on the sensitivity of the assay with the optimal (minimal) antibody titer and decreased concentrations of coating antigen. The optimal concentration of coating antigen, 0.1 ug/mL, permits the use of a mixture of coating antigens without exceeding the binding capacity in the well. There are five different coating antigens/well. Four are specific for 4 different SUs and one is an analogue which captures the spécifie antibody from five different SUs. Table Π. Coating Antigen Optimization % Inhibition* [Coating Antigen] 5 ppt Wppt 25 ppt

0.05^g/mL O.^g/mL 0.2 Ug/mL

35 33 13

53 50 21

70 68 27

50ppt

78 76 30

*% Inhibition = ((A-B)/A) χ 100 where A = Absorbance of Negative Control Β = Absorbance of Sample Incubation time is always optimized for the specific assay and specific matrix. The goal is to establish conditions that will result in a reproducible and robust assay, day to day and lab to lab. Once incubation times have been established, they are strictly adhered to. Under assay conditions, 30 minute incubation at each step is optimal. Decreasing or increasing incubation times will alter the final development time and may change the % inhibition on the curve since this is a non-equilibrium technique. Assay development time is optimized to restore the assay time lost by using a decreased amount of antibody and coating antigen. This may be accomplished by varying the concentration of the second antibody enzyme. In Table ΙΠ the effect of increasing concentration of the second antibody enzyme on assay time may be seen. This is accomplished without changing anything in the assay except development time. Table IIL Optimization: Second Antibody Enzyme (Ab-E) Conjugate Ab-E Conjugate M /mL

1.2 2.4 4.8

155 minutes 110 minutes 80 minutes

Increasing 2nd Ab-E concentration decreases total assay time without affecting % inhibition. To validate a polyantigen coated microwell plate, a comparison was made between the specific coated plate and the polyantigen coated plate for each of the nine specific SU assays. The % inhibition of each standard in each SU assay showed nearly identical results on the two plates (Table IV). Slight variations are accounted for by the controls that are included on every single plate.

Van Emon et al.; Environmental Immunochemical Methods ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

7. STRAHAN

ELISA Method for Analysis of Multiple SUs Table IV. Antigen Vs. Polyantigen Microtiter Plates % Inhibition* Polyantigen Antigen SU ppt Plate Plate

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

15.6 31.2 62.5 125 500

20 34 60 78 91

16 30 55 75 89

*% Inhibition = ((A-B)/A) χ 100 where A = Absorbance of Negative Control Β = Absorbance of Sample Benlate Investigation

Benlate 50 DF is a fungicide formulation which contains 50% Benomyl, the active ingredient and 50% inert ingredients. The DF stands for dry flowable. An investigation of 1313 lots of Benlate 50 DF was done to determine if there were any SUs present in the formulated product. These were all the lots remaining in DuPont's possession and included lots manufactured between 1987 and 1991. In addition, 216 duplicate boxes were also analyzed. Nine DuPont SUs were in the screening assay. The limit of detection (LOD) was 5 parts per billion (ppb) in the dry formulation in the box. Polyantigen coated microwell plates were ideal for this large study because the same microwell plate could be used with any antibody. Therefore, any of the nine antibodies in the study could be assayed without preparing different types of microwell plates coated with the spécifie coating antigen. The most efficient plate format is custom designed to maximize sample throughput and minimizetimeand cost. The format chosen is shown in Figure 2 and shows 3 sets of controls and 30 samples per plate. Logistically, one antibody on a plate worked best. Alternatively, fewer samples/plate could be analyzed with all nine antibodies (Figure 3) but found this approach not as efficient with the large number of samples in this study. The aqueous extraction protocol is shown in Table V. A final 1:100 dilution of the extract was made to minimize the interference from some of the inert ingredients in the formulation. With a limit of detection (LOD) of 5 ppb and this dilution, the actual measurement is 0.05 ppb in the assays or 50 parts per trillion (ppt) which is about 50% inhibition for most of the SU standard curves. A typical SU standard curve is shown in Figure 4. This (LOD) was optimal for a reliable and reproducible measurement given assay sensitivity and potential matrix interferences. Table V. Extraction Protocol

Τ

Weigh 1 gram Benlate 50 DF into a 50 mL polypropylene centrifuge tube 2. Add 20 mL phosphate buffered saline (PBS); vortex and tumble 30 minutes 3. Centrifuge 15 minutes at 5000 RPM, 0 C 4. Decant and filter supernatant 5. Dilute supernatant a further 1:5 in PBS (final dilution = 1:100) e

Van Emon et al.; Environmental Immunochemical Methods ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

69

70

E N V I R O N M E N T A L I M M U N O C H E M I C A L

1 A sb Β Ctl 0 C D CtlO M

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

Ε

F G H

CtlO

2 3 4 5 sb sb sb sb Ctl 5 Ul U2 U3 II " Ctl 5 u n U12 U13 II " Ctl 5 U21 U22 U23 »t

nsb

II

nsb

nsb

6 sb U4

7 sb U5

II

9 8 // 12 10 sb sb sb sb sb U6 U9 U10 U7 U8 " U16 U17 U18 U19 U20 II II II " U26 U27 U28 U29 U30 11

U14 " U24

U15 " U25

II

·*

nsb

nsb

M E T H O D S

nsb

nsb

nsb

nsb

nsb

Coating Antigen

All 96 wells

Control (Ctl) 0: Benlate 50 DF lot #U072390-713 Control (Ctl) +: Benlate 50 DF lot # U072390-713 + 5 ppb of a sulfonylurea

Column 1 Column 2

Benlate 50 DF Samples (U1-U30)

Columns 3-12

Substrate Blank (sb) Non-specific Binding Blank (nsb)

Row A Row H

nsb

nsb

Figure 2. Microwell Plate Format Used in ELISA Screening Method for the Detection of Nine DuPont Sulfonylurea Herbicides in 1313 Lots of Benlate 50 DF Fungicide

I 3 Ab 1 0 0 S S Ul Ul u U2

4 1 5 Ab2 0 0 S s Ul Ul U u

0 S Ul U2

Ab6 0 0 S S Ul Ul U2 U2

Ab7 0 0 S s Ul Ul U2 U2

Ab8 0 0 S S Ul Ul U2 U2

2

A Β C D Ε

F G H

2

2

2

7

6 A b3

0 S Ul U2

9

8 A b4 0 S Ul U2

0 S Ul U2

// 10 A b5 0 0 S S Ul Ul U2 U2

12

Ab9 0 0 S S Ul Ul u U2 2

Microwell Plate Format 0 = negative control S = Spiked control U = Unknown sample Figure 3. Microwell Plate Format Used in ELISA Screening Method for Simultaneous Analysis of Nine DuPont Sulfonylureas

Van Emon et al.; Environmental Immunochemical Methods ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

7. STRAHAN

ELISA Method for Analysis of Multiple SUs

71

Curve Fit: 4 Parameter C o m CoetY: 0.992 Equation: y=(A-D)/(l + (x/C) B) + D A = 1.94 Β = 1.47 C = 0.0234 D = 0.0009

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

A

ο

0J 0.0

Η

1

h-

Log scale of ppt

I— 1E+03

ng/mL

Mean AU (n=3)

% CV(AU)

0.00 0.025 0.050 0.100 0.200

1.964 1.522 1.204 0.868 0.603

1.4 1.6 2.5 2.1 2.0

Figure 4. Typical Standard Curve

Table VI shows typical recovery of each SU spiked into the Benlate 50 DF formulation at 5 ppb and then extracted. Recoveries were measured in quantitative assays and averaged about 90% but rangedfrom80% to 100%. Table VI. % Recovery of 5 ppb of each SU Benlate 50 DF (Unspiked and Spiked Measurements in ppb) Sulfonylurea % Recovery Unspiked Spiked

Nicosulfuron Metsulfuron Chlorimuron Tribenuron Chlorsulfuron Thifensulfuron Bensulfuron Ethametsulfuron Sulfometuron

0.0 1.1 0.0 0.0 0.0 0.1 0.0 0.2 0.6

4.9 5.1 4.6 4.2 5.0 5.0 5.0 4.5 4.6

98 80 92 84 100 98 100 86 80

Van Emon et al.; Environmental Immunochemical Methods ACS Symposium Series; American Chemical Society: Washington, DC, 1996.

72

ENVIRONMENTAL IMMUNOCHEMICAL METHODS

The controls were unspiked Benlate for the negative control and spiked Benlate at 5 ppb with each of the SUs for the positive controls. These sample were extracted and treated exactly the same as the test lots.

Downloaded by UNIV OF TEXAS AT AUSTIN on April 16, 2018 | https://pubs.acs.org Publication Date: October 23, 1996 | doi: 10.1021/bk-1996-0646.ch007

Assay Results

Table VH is a summary of the results. Samplesfrom1529 boxes, (1313 discrete lots plus an additional 216 duplicate boxes) were analyzed. Screening for 9 SUs in a box then represented 13,761 assays plus 157 X 9 controls (= 1413) or a total of 15,174 assays. Use of the polyantigen plate and the quantitative screening format significantly reduced thetimeand cost of this study. One technician was able to screen as many as 300 individual samples a day plus controls. The control and sample populations shown in Table V u are clearly distinct The control population is the Benlate 50 DF fortified at 5 ppb with each of the nine sulfonylureas compared to the unfortified Benlate 50 DF. % Inhibition is calculated vs the unfortified Benlate 50 DF. The sample populations are the 1313 discrete lots plus some duplicates. These are extracted and assayed along with the controls and % Inhibition is calculated vs the unfortified Benlate 50 DF. The criteria of three standard deviations from the mean shows no overlap except for Sulfometuron where the control and standard populations are separated by two standard deviations. There was excellent reproducibility of the fortified controls. This is a tribute to the skill of the technicians performing the assay and the robustness of the ELISA assay.

Table VII. % Inhibition (%I) of the Positive Controls and Benlate 50 DF Fungicide Lots Samples Positive Control